Using Geographical Weighting and Knowledge Graph Centrality to Identify Key Management Areas for Shared Bikes

Owing to the increasing demand for improving the utilization rate of shared bikes, identifying their key management areas is necessary for shared‐bike companies to effectively allocate resources and formulate efficient management strategies. However, traditional methods often assist management decisions by analyzing the spatial distribution characteristics of each influencing factor of shared‐bike usage and fail to quantitatively consider the comprehensive impact of influencing factors on the management area. Therefore, a new method for identifying key management areas for shared bikes using geographical weighting and knowledge graph centrality is proposed. In this study, a multiscale geographically weighted Poisson regression (MGWPR) model was initially used to explore the influencing factors of shared‐bike usage and their degrees of influence. The regression model results were linked to geographical statistical units to construct a knowledge graph of factors affecting shared‐bike usage in each district, and the weighted degree centrality classification results of nodes in each district were used to assist in identifying their key management areas. The results showed that the proposed method could quantitatively measure the comprehensive impact of different factors in each district on the utilization rate of shared bikes and could effectively identify their key management areas, thereby assisting enterprises in formulating appropriate shared‐bike management strategies and improving the efficiency of shared‐bike management decision‐making.